effects of left bundle-branch block on cardiac structure, function, perfusion, and perfusion...

Effects of Left Bundle-Branch Block on Cardiac Structure,Function, Perfusion, and Perfusion Reserve

Implications for Myocardial Contrast Echocardiography VersusRadionuclide Perfusion Imaging for the Detection of Coronary

Artery Disease

S.A. Hayat, MBChB, MRCP; G. Dwivedi, MD, DM, MRCP; A. Jacobsen, MD; T.K. Lim, MD;C. Kinsey, HND; R. Senior, MD, DM, FRCP, FESC

Background—We aimed to investigate the cardiac effects of left bundle-branch block (LBBB) using myocardial contrastechocardiography (MCE) to ascertain the value of MCE for detecting coronary artery disease (CAD) and to uncover themechanism that affects the accuracy of single-photon emission computed tomography (SPECT) in these patients.

Methods and Results—Sixty-three symptomatic LBBB patients (group A), 10 left ventricular ejection fraction–matchedcontrol subjects without LBBB and no CAD (group B), and 10 normal control subjects (group C) underwent restingechocardiography. Rest and vasodilator MCE and SPECT were undertaken in LBBB patients. Septal (SW) and posteriorwall (PW) thickness, thickening, quantitative myocardial blood flow (MBF), and MBF reserve were measured. SW/PWthickness and percentage thickening ratios were lower (P�0.01 and P�0.05, respectively) in group A compared withboth groups B and C, but resting SW/PW MBF and MBF reserve ratios were similar in all 3 groups. MBF reserve butnot MBF was reduced in groups A and B (2.2�0.7 versus 2.2�0.2; P�0.98) compared with group C (3.1�0.5;P�0.01). SW thickness was an independent predictor (P�0.006) of SPECT perfusion defects in LBBB patients withoutCAD. MCE (92%) had a sensitivity similar to SPECT (92%); however, the specificity of MCE (95%) was superior(P�0.0001) to SPECT (47%) for the detection of CAD.

Conclusions—Despite asymmetrical reduction in SW thickness and function, MBF is preserved and MBF reserve ishomogeneously reduced in LBBB patients with left ventricular systolic dysfunction. Because of partial volume effects,the accuracy of SPECT for detecting CAD was significantly compromised compared with MCE in this patient cohort.(Circulation. 2008;117:1832-1841.)

Key Words: bundle-branch block � coronary disease � echocardiography � nuclear medicine � perfusion

Complete left bundle-branch block (LBBB) is a commonECG disorder that often is associated with coronary

artery disease (CAD).1 LBBB in the presence of CAD isassociated with a 3- to 4-fold increase in cumulative cardio-vascular mortality.2 The accuracy of commonly used nonin-vasive techniques such as single-photon emission computedtomography (SPECT) for the detection of CAD in thesepatients is confounded by the heterogeneous effects of LBBBon myocardial structure, function, and perfusion, resulting ina high incidence of anteroseptal and septal perfusion defectsin the absence of CAD.3–6 Alternative approaches, includingthe use of coronary vasodilators as stress agents in place ofexercise or dobutamine and the application of different imageinterpretation algorithms, have reduced but not eliminated theincidence of false-positive results.6,7 Several hypotheses have

been postulated to explain the cause of such perfusion defectsseen on SPECT, including early activation of the septum,leading to shortened diastole and hence reduced blood flow8;partial volume effects caused by septal thinning and impair-ment of thickening noted in experimental models of LBBBwith right ventricle pacing9; and increased septal intramyo-cardial pressure during diastole, resulting in reduced flowreserve.10

Clinical Perspective p 1841

Quantitative myocardial contrast echocardiography(MCE), which uses contrast agents that are entirely intravas-cular, has been shown in experimental models and humans tobe accurate in assessing myocardial perfusion both at rest andduring stress.11–13 The aims of the study were to investigate

Received February 22, 2007; accepted January 22, 2008.From the Department of Cardiology and Institute of Postgraduate Medical Education and Research, Northwick Park Hospital, Harrow, UK.Correspondence to Professor Roxy Senior, MBBS, MD, DM, FRCP, FESC, FACC, Consultant Cardiologist and Director of Cardiac Research,

Honorary Professor, Middlesex University London, Honorary Senior Lecturer, Imperial College, London, Department of Cardiology, Northwick ParkHospital, Harrow HA1 3UJ, UK. E-mail [email protected]

© 2008 American Heart Association, Inc.

Circulation is available at http://circ.ahajournals.org DOI: 10.1161/CIRCULATIONAHA.107.726711

1832

Imaging

by guest on April 17, 2016http://circ.ahajournals.org/Downloaded from

http://circ.ahajournals.org/

the effects of LBBB on myocardial structure, function, andperfusion with MCE; to ascertain the value of MCE for thedetection of CAD in LBBB patients; and to test the hypoth-esis that the partial volume effects caused by septal thinningand impaired thickening rather than true hypoperfusion areresponsible for SPECT perfusion defects in patients with-out CAD.

MethodsPatient PopulationSymptomatic patients with LBBB referred for coronary arteriogra-phy by their treating physician were included in the study. Patientswith known CAD, angina, history of prior myocardial infarction,history of revascularization, significant valvular disease, or hemo-dynamic instability were excluded. Furthermore, 10 age- and leftventricular (LV) ejection fraction (LVEF) –matched control subjectswithout LBBB and with no coronary arteriographic evidence of CAD(�50% luminal diameter stenosis of any major epicardial artery ormajor branch) also were recruited. This was to account for theconfounding effects of LVEF on cardiac structure, function, andperfusion. Ten age- and sex-matched patients admitted with chestpain with nondiagnostic ECG changes, no cardiac enzyme rise, andnormal regional and global LV function both at rest and during stressechocardiography acted as normal control subjects.

The study complied with the Declaration of Helsinki. All patientsgave written informed consent to the study protocol, which wasapproved by the Institutional Review Board of Northwick ParkHospital.

Study DesignLBBB was established by a standard 12-lead ECG before studyrecruitment and was confirmed by repeat 12-lead ECG on the studyday. All patients underwent transthoracic 2-dimensional echocardi-ography, MCE, SPECT, and coronary arteriography. Stress MCEand SPECT were performed on the same day.

ECGLBBB was defined as QRS duration �120 ms, the presence ofnotched R waves in the lateral precordial leads (V5 and V6) and leadsI and aVL, small or absent initial r waves in the right precordial leads(V1 and V2) followed by deep S waves, absent septal q waves inleft-sided leads, and a prolonged intrinsicoid deflection (�60 ms) inV5 and V6.

Two-Dimensional EchocardiographyTwo-dimensional echocardiography was performed in standard api-cal and parasternal views with tissue harmonic imaging (SONOS7500, Philips Medical Systems, Best, the Netherlands). LVEF, LVend-diastolic volume, and LV end-systolic volume were measuredwith Simpson’s apical biplane rule.14 End diastole was identified bythe onset of the QRS complex; end systole was defined as thesmallest LV diameter. Tracing of the LV contour excluded thepapillary muscles and trabeculations within the LV cavity as recom-mended by the American Society of Echocardiography.14 Regionalwall thickness (WT) for the septal wall (SW) and posterior wall(PW) was measured at 3 points in the cardiac cycle (end diastole,midsystole [midpoint between the onset of QRS and end systole],and end systole) from the parasternal long-axis 2-dimensional view;maximum WT was assessed; and a mean value was obtained. Anarrow sector was used to maximize frame rate (65 to 75 Hz).Percentage of SW and PW thickening were calculated using thefollowing formula: ([maximal WT�end-diastolic WT]�100)/end-diastolic WT. Ratios of SWT to PWT and percentage thickening alsowere calculated.

MCE StudiesMCE was performed with SonoVue (Bracco Research SA, Geneva,Switzerland) at rest in the 3 apical views (apical 4-chamber,

2-chamber, and 3-chamber views) using low-power MCE (powermodulation technique) at a mechanical index of 0.1. Backgroundgains were set so that minimal tissue signal was seen. The color gainswere then adjusted so that no Doppler signal was seen except at themitral valve and proximal to the apex. SonoVue was infused at 50 to70 mL/h with VueJect (BR-INF 100, Bracco Research SA), aninfusion syringe pump that rotates gently throughout infusion tomaintain microbubble suspension. The infusion rate was adjusted toobtain the best possible myocardial opacification with minimalattenuation. Once optimized, the machine settings were kept constantthroughout the study. The focus was set at the mitral valve level butmoved toward the apex if there was concern about a near-fieldartifact. Nonstandard apical views (eg, bringing the lateral wall intothe sector field) were used, if required, to attempt to overcome basalattenuation artifacts. Flash echocardiography at a high mechanicalindex (1.0) was performed to achieve myocardial microbubbledestruction, after which 15 consecutive end-systolic frames weredigitally captured in each view (3 sequences in each) at a lowmechanical index (0.1). Imaging was repeated at peak hyperemiaafter administration of 0.56 mg/kg dipyridamole.

Normal myocardial perfusion at rest was considered to be presentwhen all segments (16-segment model) showed homogeneous con-trast opacification within 5 seconds after destructive imaging. Areversible perfusion defect was considered present when a contrastdefect was observed 1 second after destructive imaging after vaso-dilator stress with the presence of a transmural defect filling from thesubepicardium to subendocardium during subsequent cardiac cy-cles.15 CAD was considered to be present when 2 contiguoussegments demonstrated perfusion defects (rest or stress). Multivesseldisease was considered present when perfusion defects were noted in�1 vascular territory. The analysis was performed by an investigator(R.S.) who was blinded to clinical, coronary arteriographic, andSPECT data.

Quantitative MCEQuantitative MCE was performed offline by a single observer (S.H.)unaware of the 2-dimensional echocardiographic or angiographicdata. Standard commercial software (Q-Laboratory, Philips MedicalSystems) was used to quantify myocardial replenishment for 15frames after bubble destruction by placing the region of interestacross the entire thickness of the myocardium, excluding the high-intensity endocardial and epicardial borders.2 The basal segments ofthe 16-segment LV model were not included in the quantitativeanalysis. QLab software automatically constructed background-subtracted plots of peak myocardial contrast intensity, representingmyocardial blood volume (A), versus pulsing intervals, from whichthe slope of the replenishment curve depicting mean microbubblevelocity (�) was derived. Frames showing wide variation in contrastintensity were discarded to minimize errors in the analysis. Myocar-dial blood volume was subsequently normalized for blood poolvideointensity by placing the ventricular cavity region of interestadjacent to the myocardial region of interest using a previouslyvalidated equation: normalized A�10([A�blood pool]/10)�100.13 SW andPW myocardial blood flow (MBF) and MBF reserve (MBFR) werecalculated from these formulas: MBF�A�� and MBFR�stressMBF/resting MBF. Ratios of SW to PW MBF and MBFR also werecalculated.

Gated Technetium-99m Sestamibi SPECTA separate-day stress-rest protocol was used. A rest study wasperformed on a separate day 60 to 90 minutes after injection of 600MBq technetium-99m sestamibi. For stress imaging, 600 MBqtechnetium-99m sestamibi was injected at peak hyperemia afteradministration of dipyridamole (during stress MCE). Images wereacquired 60 to 90 minutes after injection with a multihead camerawith high-resolution collimators. The camera energy window (20%)was set on the 140-keV photopeak of technetium-99m sestamibi.Particular care was taken to avoid patient motion and overlap fromextracardiac activity. A total of 32 projections (each lasting 40seconds) were acquired over a 180° arc from the 45° right anterioroblique to the 45° left posterior positions. Both rest and stress

Hayat et al Left Bundle-Branch Block and MCE 1833



tomograms were reconstructed in the vertical and horizontal long-and short-axis planes and were simultaneously analyzed with astandard 16-segment LV model. SPECT perfusion was graded witha well-validated scoring system: 0�normal tracer uptake, 1�mildlyreduced tracer uptake, 2�moderately reduced tracer uptake, 3�severely reduced tracer uptake, and 4�absent tracer uptake.16

A rest defect was defined as a score of �2 in at least 1 segment(matching 16-segment LV model). A fixed defect was any restingdefect that remained unchanged during stress with an accompanyingwall thickening abnormality on gated SPECT. A reversible defectwas defined as a reduction in tracer uptake by at least 1 grade exceptwhen the resting score was 0 when the change in score should be �2.CAD was considered present when a resting defect and/or areversible defect were detected in �2 contiguous segments. SPECTmyocardial quantification was performed with MyoQuant soft-ware.17 The software calculates and quantifies perfusion and perfu-sion deficits in myocardial SPECT data through the analysis of polarmaps generated from the radial slices. Normalized perfusion valueswere displayed in the 16-segment grid model. The qualitative andquantitative assessments of SPECT data were performed by 2independent observers (A.J. and C.K.) who were blinded to clinical,MCE, and coronary arteriographic data.

Coronary ArteriographySelective coronary arteriography was performed, and patients withCAD were defined as those demonstrating �50% luminal diameterstenosis of any major epicardial artery or major branch. Multivesseldisease was defined as CAD in the left anterior descending artery andright coronary artery or left circumflex artery.

Statistical AnalysisResults from normally distributed continuous data are expressed asmean�SD. Categorical variables are presented as percentages. Thepaired t test was used to compare the differences within groups, andthe independent t test was used to compare continuous variablesbetween groups. One-way ANOVA was used with post hoc Tukeyadjustment in the multiple comparisons between different groups ofcontinuous variables to minimize the type I error. We used �2 teststo compare categorical variables between groups. A receiver-operating characteristics curve was plotted to determine the bestMBF reserve cutoff value for predicting significant CAD. One-wayANOVA was used to compare the MCE variables among variousgrades of stenosis. The presence of abnormality in either of thecoronary territories was considered to be positive for the detection ofCAD on a per-patient basis.

The effect of various parameters for the prediction of SPECTperfusion defect was obtained by logistic regression analysis. Sub-sequently, the joint effects of the explanatory factors on the SPECTperfusion defect were examined together in a multivariate analysis.Only factors with values of P�0.1 in the univariable analysis wereincluded in the multivariable analysis. A backwards selection pro-cedure was used to determine the final model. McNemar’s test wasused to compare the sensitivity and specificity of MCE and SPECT.A value of P�0.05 (2 sided) was considered significant. Statisticalanalysis was performed with Analyze-it software for Microsoft Excel(version 1.62, Analyze-it Software Ltd, Leeds, UK) and SPSSversion 14.0 (SPSS Inc, Chicago, Ill).

The authors had full access to and take full responsibility for theintegrity of the data. All authors have read and agree to themanuscript as written.

Results

Patient DemographicsThe demographics of the LBBB patient population aresummarized in Table 1. Table 2 showed no significantdifferences in age and Framingham risk score between the 3groups.

Correlates of Myocardial Structure in LBBBThe ratio of SWT to PWT was significantly lower in group A(LBBB and no CAD) compared with groups B (LVEF-matched control subjects without LBBB and no CAD) and C(normal control subjects), whereas no differences were notedbetween groups B and C (Table 3). This was because SWT ingroup A was significantly lower than in both groups B and C.

Correlates of Myocardial Function in LBBBThe ratio of percentage of SW to PW thickening wassignificantly lower in patients with LBBB (group A) com-pared with groups B and C, whereas no differences werenoted between groups B and C (Table 3). The primary reasonfor the reduced ratio of SW to PW thickening in group A wasthat SW thickening was lower compared with group B despitesimilar reductions in LVEF and comparable LV volumes.Both groups A and B showed significant reductions inpercentage SW and PW thickening compared with normalcontrol subjects (Group C).

Correlates of Myocardial Perfusion in LBBBAlthough LBBB patients demonstrated smaller SWT andpercentage wall thickening compared with groups B and C,there were no significant differences in indexes of MBF(Table 3). Resting A, �, and MBF were similar across all 3groups, as were the ratios of SW to PW MBF (1.01�0.14,0.99�0.15, and 1.02�0.15, respectively). Similarly, no sig-nificant differences were noted in patients with LBBB with

Table 1. Patient Characteristics

No CAD CAD P

Patient demographics, n (%) 38 (60) 25 (40)

Age, y 67�9 71�10 0.10

Men, n (%) 19 (50) 16 (64) 0.40

Hypertension, n (%) 18 (47) 12 (48) 0.83

Diabetes, n (%) 10 (26) 9 (36) 0.59

Hyperlipidemia, n (%) 12 (32) 10 (40) 0.68

Smoking history, n (%) 6 (16) 5 (20) 0.93

Family history of CAD, n (%) 5 (14) 5 (20) 0.71

Framingham risk score, % 15�7 17�9 0.37

Presenting symptoms, n (%)

Chest pain 24 (63) 16 (64) 0.84

Shortness of breath 25 (66) 20 (80) 0.35

Coronary arteriography, n (%)

1-Vessel disease � � � 12 (48) � � �

2-Vessel disease � � � 7 (28) � � �

3-Vessel disease � � � 6 (24) � � �

LAD disease � � � 18 (72) � � �

RCA/LCx disease � � � 19 (76) � � �

Multivessel disease (LAD�RCA/LCx) � � � 12 (48) � � �

LVEF, % 37�14 42�14 0.16

LAD indicates left anterior descending coronary artery; RCA, right coronaryartery; and LCx, left circumflex artery. n�63.

1834 Circulation April 8, 2008



and without CAD for A (P�0.53) and MBF (P�0.49), but �

tended to be lower in the CAD group (P�0.07). Resting MBFwas not reduced in patients with LBBB (group A) comparedwith LVEF-matched control subjects without LBBB (groupB) and normal control subjects (group C). Similarly, theratios of SW to PW MBFR were similar in all 3 groups.However, MBFR was reduced equally in groups A and B(2.2�0.7 versus 2.2�0.2; P�0.98) compared with normalcontrol subjects (3.1�0.5; P�0.0005; Figure 1).

Comparison of LBBB Patients With andWithout CADNo significant differences were noted in LBBB patients withand without CAD for SWT (P�0.60), PWT (P�0.51), andratio of SWT to PWT (P�0.37). Similarly, there were nosignificant differences for percentage of SW thickening(P�0.56), percentage of PW thickening (P�0.17), and ratioof SW to PW thickening (P�0.49). However, in the LBBBgroup with CAD, MBFR (0.78�0.46) was significantly

Table 2. Comparison Between Patients With LBBB Without CAD, No LBBB Without CAD (LVEFMatched), and Normal Control Subjects

Group A; No CAD,LBBB

Group B; No CAD,No LBBB

Group C; Normal ControlSubjects P

N 38 10 10

Mean age, y 67�9 64�9 64�6 NS

Men, n (%) 19 (50) 3 (30) 6 (60) NS

Hypertension, n (%) 18 (47) 6 (60) 5 (50) NS

Diabetes, n (%) 10 (26) 3 (30) 3(30) NS

Hyperlipidemia, n (%) 12 (32) 4 (40) 2 (20) NS

Smoking history, n (%) 6 (16) 2 (20) 2 (20) NS

Family history of CAD, n (%) 5 (14) 4 (40) 3 (30) NS

Framingham risk score, % 15�7 14�9 16�5 NS

LVEF, % 37�14* 35�8* 62�5 �0.001

LVESV, m/s 96�69† 90�30‡ 39�11 0.03

LVEDV, m/s 141�65‡ 131�39‡ 103�24 NS

LVESV indicates LV end-systolic volume; LVEDV, LV end-diastolic volume; and NS, not significant between the 3groups. The �2 test was used for categorical variables; 1-way ANOVA was used for continuous variables.

*P�0.001, †P�0.02, ‡P�NS vs group C.

Table 3. Comparison of Myocardial Structure, Function, and Perfusion in LBBB Patients (No CAD) With No LBBB, LVEF-MatchedControl Subjects, and Normal Control Subjects

Group A; LBBB,No CAD (n�38)

Group B; No LBBB, Age andLVEF Matched (n�10)

Group C; Normal Control Subjects,Age Matched (n�10) P (1-Way ANOVA)

Parameters of myocardial structure

Mean SW thickness, cm 1.22�0.17*†‡ 1.35�0.10 1.38�0.11 0.004

Mean PW thickness, cm 1.32�0.15 1.33�0.08 1.38�0.07 NS

Mean SWT/PWT ratio 0.93�0.07§� 1.02�0.07 1.00�0.07 �0.0001

Parameters of myocardial function

SW thickening, % 23.8�5.4¶#** 28.0�4.5†† 35.8�4.7 �0.001

PW thickening, % 27.4�5.0 27.5�4.7 34.9�5.4‡‡ �0.001

SW/PW thickening ratio 0.88�0.17§§�� 1.03�0.14 1.03�0.07 0.004

Parameters of myocardial perfusion

SW resting A 7.9�2.8 7.8�2.1 7.9�3.3 NS

SW resting � 0.73�0.13 0.70�0.14 0.69�0.10 NS

SW resting MBF 5.5�2.0 5.4�1.6 5.5�2.6 NS

SW MBFR 2.2�0.7¶¶ 2.2�0.2## 3.1�0.5 �0.005

SW/PW resting MBF ratio 1.01�0.14 0.99�0.15 1.02�0.15 NS

SW/PW MBFR ratio 1.07�0.43 1.05�0.15 1.06�0.16 NS

The probability values in each subsection are for pairwise comparisons using post hoc Tukey adjustment to account for multiple comparisons.*P�0.05 vs group B; †P�0.01 vs group C; ‡P�0.0001 vs PW, §P�0.005 vs group B; �P�0.002 vs group C; ¶P�0.05 vs group B; #P�0.001 vs group C;

**P�0.0003 vs PW; ††P�0.002 vs group C; ‡‡P�0.005 vs groups A and B; §§P�0.03 vs group C; ��P�0.02 vs group C; ¶¶P�0.001 vs group C; ##P�0.0001vs group C.




reduced compared with LBBB patients without CAD(P�0.0001).

Distribution of Perfusion DefectsOf the 38 patients without CAD, 20 (53%) demonstratedperfusion abnormalities on SPECT, 18 of which (90%) wereseptal. The majority of perfusion defects were fixed. MCEdemonstrated normal perfusion in all these patients; however,MCE demonstrated perfusion abnormalities in 2 patients(5%) with normal SPECT perfusion. The distribution ofperfusion abnormalities by the 2 techniques is illustrated inFigure 2.

Relation Between Regional WT, Function, andIndexes of MBF in Patients With and WithoutSPECT Perfusion DefectsQuantitative SPECT analysis revealed that the ratio of SW toPW count was significantly reduced (P�0.01) in patientswith visual SPECT perfusion defects (0.79�0.06) comparedwith those without SPECT defects (0.85�0.07). Furthermore,perfusion defects were significantly (P�0.009) larger in theseptum in patients with visual perfusion defects (26�8%)compared with those without (18�11%). The qualitativetracer uptake score in patients with perfusion defects was

2.5�0.27, with 97% of the defects classified as moderate tosevere. Mean SWT, end-diastolic SWT, SW thickening, andLVEF were significantly lower in patients with septal SPECTperfusion defects compared with those without such de-fects (Table 4). Mean SWT was significantly lower thanmean PWT (1.15�0.12 and 1.28�0.13 cm, respectively;P�0.0001) in patients with SPECT perfusion defects. How-ever, there was no difference between mean SWT and meanPWT in patients with normal SPECT perfusion (1.35�0.20and 1.35�0.21 cm, respectively; P�0.92). On multivariableanalysis, mean SWT emerged as the only independent pre-dictor (P�0.006) of false SPECT perfusion defects. The

0

0.5

1

1.5

2

2.5

3

3.5

4

A B C

GROUP

MB

FR

p <0.0005

p = 0.98

Figure 1. Comparison of MBFR between the3 groups. A, LBBB and no CAD. B, No LBBB,no CAD LVEF-matched control subjects. C,Normal control subjects.

2 (10%) Antero-apical

Total patients (n=38)

MCE: 2 (5%) patients perfusion defect

SPECT:20 (53%) patients with perfusion defects

1 Apical defect

18 (90%) Antero-septal

Fixed: 16 (80%) Reversible: 4 (20%)

1 Posterior defect

Figure 2. Distribution of perfusion defects in patients with LBBBand no CAD.

Table 4. Univariable Predictors of Septal SPECT PerfusionDefect in Patients With LBBB and No CAD

Normal Perfusion(n�18)

Perfusion Defect(n�20) P

Resting HR, bpm 70�11 70�14 0.93

Stress HR, bmp 83�11 84�14 0.82

Mean PWT, cm 1.35�0.21 1.28�0.13 0.21

Mean SWT, cm 1.35�0.20 1.15�0.12 0.001

Diastolic PWT, cm 1.18�0.17 1.11�0.11 0.14

Diastolic SWT, cm 1.12�0.16 0.99�0.10 0.005

PW thickening, % 27.6�5.9 26.9�4.2 0.66

SW thickening, % 26.4�5.9 21.4�4.1 0.004

LVEF, % 42�13 32�15 0.04

LVESV, mL 82�54 116�73 0.11

LVEDV, mL 130�60 159�70 0.17

Posterior resting MBV, dB 8.01�2.93 7.15�2.30 0.26

Posterior resting MBF, dB/s 5.6�1.95 5.3�1.68 0.65

Posterior MBFR 2.16�0.55 2.03�0.85 0.56

Septal resting MBV, dB 8.37�3.37 7.49�2.51 0.36

Septal resting MBF, dB/s 5.7�2.32 5.5�2.21 0.78

Septal MBFR 2.15�0.72 2.30�0.88 0.59

HR indicates heart rate; LVESV, LV end-systolic volume; LVEDV, LVend-diastolic volume; and MBV, myocardial blood volume.




end-diastolic WT cutoff of 1.02 cm provided a sensitivity of80% and specificity of 83% (area under the curve, 0.88;Figure 3). Figure 4 shows an example of a patient withnormal coronary arteriography who had a normal MCE studybut whose SPECT revealed significant fixed perfusion de-fects in the septum and apex.

Diagnostic Accuracy of Quantitative MCE for theDetection of CADSensitivity and specificity for the detection of CAD (�50%luminal diameter stenosis) for 63 anterior and 62 posterior(failure to intubate right coronary artery in 1 patient) coronaryterritories by quantitative MCE were determined by plottinga receiver-operating characteristics curve (Figure 5). The areaunder the curve was higher for the detection of CAD in theanterior circulation (0.95) compared with posterior circula-

tion (0.88). An MBFR cutoff of 0.85 provided a sensitivity of89%, specificity of 91%, positive predictive value of 80%,and negative predictive value of 95% for anterior circulation;those values for posterior circulation were 79%, 84%, 68%,and 90% and on a per-patient basis were 80%, 79%, 71% and86%, respectively. MBFR was significantly higher in patientswith coronary artery stenosis �50% stenosis compared withpatients with �50% stenosis (2.00�0.77 versus 0.78�0.46;P�0.0001; Figure 6). Among patients with CAD, MBFR alsowas able to differentiate between differing grades of coronaryartery stenosis (Figure 7).

Diagnostic Accuracy of Qualitative MCE for theDetection of CAD: Comparison WithQualitative SPECTOf the 25 patients with CAD, both qualitative MCE andqualitative SPECT detected CAD in 23 patients (92%).However, of the 38 patients with no CAD, MCE correctlypredicted the absence of CAD in 36 patients (95%) andSPECT in 18 patients (47%; P�0.0001). The positive pre-dictive values of MCE and SPECT were 92% and 54% andthe negative predictive values were 95% and 90%, respec-tively, for the prediction of CAD. MCE was more accuratethan SPECT for the detection of CAD (94% versus 65%;P�0.0003). At the vascular territory level, MCE correctlylocalized left anterior descending artery disease in 16 of 18(89%) compared with SPECT, which detected 13 of 18 (72%)(P�NS). The specificities of MCE and SPECT in thisterritory were 90% and 45% (P�0.001). Localization of rightcoronary artery/left circumflex artery disease was similarbetween MCE and SPECT (sensitivity, 72% versus 68%,P�NS; specificity, 80% versus 75%, P�NS). MCE wasabnormal in all 12 patients (100%) with multivessel diseaseand correctly detected 11 of 12 patients (92%) with multives-sel disease. The corresponding numbers for SPECT were83% and 58%. The number of segments demonstratingreversible defects in patients with CAD was significantly

Figure 3. Receiver-operating characteristics curve demonstrat-ing the relationship between end-diastolic SWT and the pres-ence of septal perfusion defect in patients without CAD andLBBB.

Figure 4. MCE shows normal perfusionat rest (top left) and at stress (top right).SPECT (bottom) demonstrated a fixedperfusion defect effecting the septumand apex in a patient with no CAD.




(P�0.008) greater with MCE (6.8�3.0) compared withSPECT (4.4�3.7). If only reversible defects were consideredto represent CAD, then the sensitivity of MCE on a vascularterritory basis was 84%, but the sensitivity of SPECT droppedto 57% (P�0.03).

DiscussionThis is the first study that simultaneously assessed cardiacstructure, function, perfusion, and MBFR with quantitativeMCE in patients with symptomatic LBBB who also under-went coronary arteriography. The study demonstrated that40% of such patients had CAD. Coronary risk factors andmodes of presentation were similar in patients with andwithout CAD. LVEF is reduced in these patients, but therewas no significant difference between LVEF in patients withand without CAD. In patients with LBBB, regardless ofCAD, there were asymmetrical reductions in SWT andfunction compared with the PW. However, patients withsimilarly reduced LVEF without LBBB and no CAD did notdemonstrate asymmetrical reductions in SWT and functioncompared with the PW. SWT and function in patients withLBBB but not in those without LBBB also were asymmetri-

cally reduced compared with normal control subjects. Despiteasymmetrical reductions in SWT and function in patientswith LBBB, resting myocardial blood volume and MBF werehomogeneously preserved, but MBFR was reduced, albeithomogeneously, compared with normal controls. The presentstudy further showed that SPECT perfusion defects werecommon and located predominantly within the septum inpatients with LBBB without CAD. They were largely fixedand due to partial volume effects secondary to septal thinningand reduced septal thickening. Preservation of resting myo-cardial blood volume, MBF, and homogeneous MBFR sug-gests that true hypoperfusion was not the cause of theperfusion defects seen on SPECT.

Partial volume effects are of particular relevance in myo-cardial SPECT studies because of the limited spatial resolu-tion of gamma cameras (10 to 12 mm). Our study showed thatan end-diastolic SWT of �10 mm produced a perfusiondefect in 80% of patients. Furthermore, during SPECT imageacquisition, systolic thickening and diastolic relaxation of themyocardium result in continuous changes in the recoverycoefficient because the myocardial counts actually measureddepend considerably on systolic thickening. Myocardialthickening leads to a higher recovery coefficient and subse-quently to greater myocardial counts. In our study, mean

Figure 5. Receiver-operating characteristics curve demonstrat-ing diagnostic accuracy of quantitative MCE for the detection ofCAD in patients with LBBB. A, Anterior circulation. B, Posteriorcirculation. AUC indicates area under the curve; Sens., sensitiv-ity; Spec., specificity; PPV, positive predictive value; and NPV,negative predictive value.

Figure 6. Relationship between quantitative MCE-derived MBFRand detection of significant CAD.

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

50-69 70-89 >90

Grade of Coronary Artery Stenosis (%)

MB

Fre

se

rve

p = 0.005

Figure 7. Relationship between quantitative MCE-derived MBFRand detection of differing grades of coronary stenosis.




systolic SWT was the only independent factor accounting forperfusion defects in the septum among other confoundingvariables.

Experimental models in which right ventricular pacing wasused to mimic LBBB demonstrated similar asymmetricalreduction in SWT and function. Several investigators foundthat during right ventricular pacing, regional shortening of theseptum, which is activated early, is reduced compared withthe PW, which is activated late.9,18 It is likely that as a resultof this asynchronous activation of the left ventricle, redistri-bution of mechanical load occurs in the left ventricle, somechanical load is least in the early activated SW.18 Thisprobably results in an adaptive reduction in wall thicknessand function of the septum. However, resting myocardialblood volume and perfusion are preserved despite a reductionin WT and function. Preservation of septal myocardialperfusion in patients with LBBB also was shown in recentpositron emission tomography studies using 150� water and13N-NH3, which are robust markers of MBF, as blood flowtracers.19,20 In addition, our study demonstrated that MBFR isreduced, although homogeneously, throughout the LV. Thesefindings support the hypothesis that asymmetrical reductionsin SWT and function with preserved myocardial perfusionoccur as a physiological response to a regional decrease inmyocardial load. Symmetrical reduction in MBFR probablyresulted from underlying cardiomyopathy in the absenceof CAD.

Clinical ImplicationsCardiac resynchronization therapy is now a recognized man-agement strategy in heart failure and LBBB. One of theunderlying mechanisms of the success of cardiac resynchro-nization therapy is that it likely restores synchronous activa-tion, allowing more homogeneous distribution of mechanicalload throughout the LV and thus reversing septal dysfunction.In our study, in which no patients had prior acute myocardialinfarction, myocardial viability was maintained as predictedby preserved myocardial blood volume and MBF. However,myocardial blood velocity, which reflects MBF, tended to belower in patients with CAD compared with those withoutCAD, which is consistent with hibernating myocardium.21

The presence of myocardial viability has been shown to beone of the important predictors of success after cardiacresynchronization therapy.22

We also demonstrated that despite changes in regionalmyocardial structure and function, MCE was able to detectCAD accurately. The reason is that MCE has excellent spatial(2 to 4 mm) and temporal resolution.12,13,15 MBFR wasmarkedly reduced in patients with LBBB and CAD comparedwith those without LBBB and CAD. MCE could excludenearly all patients (�90%) without CAD and detected allpatients with multivessel disease. We also have shown that asa result of partial volume effects, the specificity of SPECT fordetecting CAD was significantly reduced compared withMCE. Our study, similar to other studies, found SPECT tohave lower accuracy for predicting CAD in patients withLBBB because of a high percentage of false-positive antero-septal and septal perfusion defects.3–7 This also was shown in

a smaller study comparing MCE and SPECT.23 No differencein sensitivity between the 2 techniques was noted for thedetection of CAD. However, because reversible defects(suggesting myocardial ischemia) occurred in only 57% ofCAD patients, a significant proportion of patients undergoingSPECT would inadvertently be denied revascularization. Thelow incidence of reversibility in patients with CAD is due toa low tracer count (partial volume effect) at rest and thus anindiscernible reduction in tracer uptake during stress becauseat least a 20% to 30% change in myocardial blood volume isrequired to produce a perfusion defect. On the other hand,because MCE does not suffer from partial volume effects andbecause it tracks blood flow, unlike SPECT, it demonstratedreversible perfusion defects in �85% of patients with CAD.Furthermore, the reversible defects were larger comparedwith SPECT in patients with CAD.

Comparison With Other Competing TechniquesStress echocardiography also has reduced accuracy for de-tecting CAD in patients with LBBB because it relies onchanges in regional wall thickening and motion, both ofwhich are affected even in LBBB without CAD.24 BecauseMCE assesses myocardial perfusion, it is a function-independent technique and thus is unlikely to be affected bythe unique functional effects of LBBB. This is reflected in theexcellent diagnostic accuracy of MCE for the detection ofCAD in our study. Another emerging noninvasive techniquethat can reliably assess CAD is multislice computed tomog-raphy.25 However, assessment of symptomatic patients withLBBB not only requires demonstration of CAD but also mustinclude assessment of its functional significance and evalua-tion of the presence or absence of myocardial viability.Echocardiography with MCE has the advantage in that it canprovide a comprehensive assessment of cardiac structure,cardiac function, presence or absence of flow-limiting CAD,and status of myocardial viability in a single examination inpatients presenting with symptomatic LBBB.

Study LimitationsMost patients with LBBB in the present study demonstratedLV systolic dysfunction; thus, the conclusions drawn fromthe present study pertain largely to this specific population. Inour study population, shortness of breath was the predomi-nant symptom that may explain the low mean LVEF. Pres-ence of LV dysfunction also explains the lower MBFR in thisgroup compared with normal control subjects; MBFR wasfurther reduced in LBBB patients with CAD. Thus, ourresults do not apply to patients with LBBB and normal LVsystolic function. However, in a smaller study of asymptom-atic LBBB patients who demonstrated predominantly normalLV function, the performance of MCE and SPECT for thedetection of CAD was similar to our study.23

The control group in our study was small, and multiplecomparisons were performed, which may have led to afalse-positive result. To minimize the possibility of type Ierror rate, we performed a 1-way ANOVA with post hocadjustment for multiple comparisons with the Tukey method.In our study, factors that may affect cardiac structure,




function, and perfusion are represented reasonably in thestudy versus control groups. The main variables likely toaffect cardiac structure are age, hypertension, diabetes mel-litus, and gender. In the 3 groups, age, hypertension, anddiabetes mellitus were well represented. Furthermore, anydiscrepancy was minimized by measurement of the ratios ofthe SWT to PWT. The latter specifically negates any bias thatmay have occurred with the discrepant gender distribution inthe 3 groups. The main demographic factors that may affectfunction and perfusion include age, hypertension, diabetesmellitus, hyperlipidemia, and smoking history; most of thesefactors are not discrepantly distributed. To further negate anypossible errors when categorical variables are comparedbetween relatively small groups, we assessed Framinghamrisk score, which incorporates all the above variables into acontinuous variable. Framingham risk scores were similaracross all 3 groups.

Finally, for normal control subjects, we have includedpatients presenting with chest pain but with normal cardiacenzymes and normal resting and stress function, whichclassifies them as having a low probability of significantcardiac disease. However, cardiac disease cannot be entirelyexcluded in this group.

ConclusionsDespite a reduction in SWT and function compared with thePW, myocardial perfusion at rest is preserved and homoge-neous in patients with LBBB without CAD and LV systolicdysfunction. However, MBFR is reduced, albeit homoge-neously. Because MCE is a partial volume– and function-independent technique, the accuracy of MCE for detectingCAD is not compromised compared with SPECT.

Source of FundingThis study was funded by a grant from the Cardiac Research Fund,Institute of Postgraduate Medical Education and Research, Harrow,Middlesex, UK.

DisclosuresNone.

References1. Schneider JF, Thomas HE, McNamara PM, Kannel WB. Clinical-

electrocardiographic correlates of newly acquired left bundle branchblock: the Framingham study. Am J Cardiol. 1985;55:1332–1338.

2. Eriksson P, Wilhelmsen L, Rosengren A. Bundle-branch block inmiddle-aged men: risk of complications and death over 28 years. EurHeart J. 2005;26:2300–2306.

3. Tandogan I, Yetkin E, Ileri M, Ortapamuk H, Yanik A, Cehreli S, DuruE. Diagnosis of coronary artery disease with Tl-201 SPECT in patientswith left bundle branch block: importance of alternative interpretationapproaches for left anterior descending coronary lesions. Angiology.2001;52:103–108.

4. Hirzel HO, Senn M, Nuesch K, Buettner C, Pfeiffer A, Hess OM,Krayenbuehl HP. Thallium-201 scintigraphy in complete left bundlebranch block. Am J Cardiol. 1984;53:764–769.

5. Larcos G, Gibbons RJ, Brown ML. Diagnostic accuracy of exercisethallium-201 single-photon emission computed tomography in patientswith left bundle branch block. Am J Cardiol. 1991;68:756–760.

6. Vaduganathan P, He ZX, Raghavan C, Mahmarian JJ, Verani MS.Detection of left anterior descending coronary artery stenosis in patientswith left bundle branch block: exercise, adenosine or dobutamineimaging? J Am Coll Cardiol. 1996;28:543–550.

7. Matzer L, Kiat H, Friedman JD Van Train K, Maddahi J, Berman DS. Anew approach to the assessment of tomographic thallium-201 scintigra-phy in patients with left bundle branch block. J Am Coll Cardiol. 1991;17:1309–1317.

8. Grines CL, Bashore TM, Boudoulas H, Olson S, Shafer P, Wooley CF.Functional abnormalities in isolated left bundle branch block: the effect ofinterventricular asynchrony. Circulation. 1989;79:845–853.

9. Prinzen FW, Cheriex EC, Delhaas T, van Ooosterhout MF, Arts T,Wellens HJ, Reneman RS. Asymmetric thickness of the left ventricularwall resulting from asynchronous electric activation: a study in dogs withventricular pacing and in patients with left bundle branch block. AmHeart J. 1995;130:1045–1053.

10. Ono S, Nohara R, Kambara H, Okuda K, Kawai C. Regional myocardialperfusion and glucose metabolism in experimental left bundle branchblock. Circulation. 1992;85:1125–1131.

11. Wei K, Jayaweera AR, Firoozan S, Linka A, Skyba DM, Kaul S. Quan-tification of myocardial blood flow with ultrasound-induced destructionof microbubbles administered as a constant venous infusion. Circulation.1998;97:473–483.

12. Senior R, Janardhanan R, Jeetley P, Burden L. Myocardial contrastechocardiography for distinguishing ischemic from nonischemicfirst-onset acute heart failure: insights into the mechanism of acute heartfailure. Circulation. 2005;112:1587–1593.

13. Peltier M, Vancraeynest D, Pasquet A, Ay T, Roelants V, D’hondt AM,Melin JA, Vanoverschelde JL. Assessment of the physiologic signif-icance of coronary disease with dipyridamole real-time myocardialcontrast echocardiography: comparison with technetium-99m sestamibisingle-photon emission computed tomography and quantitative coronaryangiography. J Am Coll Cardiol. 2004;43:257–264.

14. Schiller NB, Shah PM, Crawford M, DeMaria A, Devereux R,Feigenbaum H, Gutgessell H, Reicheck N, Sahn D, Schnittger I, Sil-verman NH, Tajik AJ. Recommendations for quantitation of the leftventricle by two-dimensional echocardiography. J Am Soc Echocardiogr.1989;2:358–367.

15. Senior R, Lepper W, Pasquet A, Chung G, Hoffman R, VanoverscheldeJL, Cerqueira M, Kaul S. Myocardial perfusion assessment in patientswith medium probability of coronary artery disease and no prior myo-cardial infarction: comparison of myocardial contrast echocardiographywith 99mTc single-photon emission computed tomography. Am Heart J.2004;147:1100–1105.

16. Senior R, Kaul S, Raval U, Lahiri A. Impact of revascularization andmyocardial viability determined by nitrate-enhanced Tc-99m sestamibiand Tl-201 imaging on mortality and functional outcome in ischemiccardiomyopathy. J Nucl Cardiol. 2002;9:454–462.

17. Vivegnis BTD, Foulon J, Rigo P. Quantitative evaluation of myocardialsingle-photon emission tomographic imaging: application to the mea-surement of perfusion defect size and severity. Eur J Nucl Med. 1996;23:1603–1612.

18. Prinzen FW, Augustijn CH, Arts T, Allessie MA, Reneman RS. Redis-tribution of myocardial fiber strain and blood flow by asynchronousactivation. Am J Physiol. 1990;259:H300–H308.

19. Nowak B, Stellbrink C, Schaefer WM, Sinha AM, Breithardt OA,Kaiser HJ, Reinartz P, Hanrath P, Buell U. Comparison of regionalmyocardial blood flow and perfusion in dilated cardiomyopathy andleft bundle branch block: role of wall thickening. J Nucl Med. 2004;45:414 – 418.

20. Zanco P, Desideri A, Mobilia G, Cargnel S, Milan E, Celegon L, Buch-berger R, Ferlin G. Effects of left bundle branch block on myocardialFDG PET in patients without significant coronary artery stenoses. J NuclMed. 2000;41:973–977.

21. Selvanayagam JB, Jerosch-Herold M, Porto I, Sheridan D, ChengASH, Petersen SE, Searle N, Channon KM, Banning AP, Neubauer S.Resting myocardial blood flow is impaired in hibernating myocardi-um. Circulation. 2005;112:3289 –3296.

22. Ypenburg C, Schalij MJ, Bleeker GB, Steendijk P, Boersma E, Dibbets-Schneider P, Stokkel MP, van der Wall EE, Bax JJ. Impact of viabilityand scar tissue on response to cardiac resynchronization therapy inischaemic heart failure patients. Eur Heart J. 2007;28:33–41.

23. Karavidas AI, Matsakas EP, Lazaros GA, Brestas PS, Avramidis DA,Vacharoulis AA, Fotiadis IN, Lprres DA, Zacharoulis AA. Comparisonof myocardial contrast echocardiography with SPECT in the evaluation ofcoronary artery disease in asymptomatic patients with LBBB. IntJ Cardiol. 2006;112:334–340.




24. Geleinse ML, Vigna C, Kasprzak JD, Rambaldi R, Salvatori MP, ElhendyA, Cornel JH, Fioretti PM, Roelandt JR. Usefulness and limitations ofdobutamine-atropine stress echocardiography for the diagnosis ofcoronary artery disease in patients with left bundle branch block: amulticentre study. Eur Heart J. 2000;21:1666–1673.

25. Ghostine S, Caussin C, Daoud B, Habis M, Perrier E, Pesenti-Rossi D,Sigal-Cinqualbre A, Angel CY, Lancelin B, Capderou A, Paul JF. Non-invasive detection of coronary artery disease in patients with left bundlebranch block using 64 slice computed tomography. J Am Coll Cardiol.2006;48:1925–1934.

CLINICAL PERSPECTIVEThe present study is the first to simultaneously assess cardiac structure, function, perfusion, and perfusion reserve usingquantitative myocardial contrast echocardiography in patients with symptomatic left bundle-branch block who predomi-nantly had left ventricular dysfunction. In these patients, regardless of coronary artery disease (CAD), there wereasymmetrical reductions in septal wall thickness and function compared with the posterior wall; however, restingmyocardial blood volume and flow were homogeneously preserved. Single-photon emission computed tomography(SPECT) is widely used for the detection and risk stratification of CAD. However, we demonstrated that because ofasymmetrical reductions in septal thickness and function, SPECT has a significantly higher incidence of false-positiveperfusion defects resulting from partial volume effects as a consequence of its poorer spatial and temporal resolutioncompared with myocardial contrast echocardiography. Despite no difference in sensitivity between the 2 techniques for thedetection of CAD, because of partial volume effects, reversible defects (suggesting myocardial ischemia) occurred in only57% of CAD patients with SPECT imaging compared with 92% with myocardial contrast echocardiography. This findinghas both cost and safety implications in that patients would be inappropriately referred for coronary arteriography afterSPECT and a significant proportion of patients undergoing SPECT would inadvertently be denied revascularization.Compared with other imaging techniques such as multislice computed tomography, SPECT, or stress echocardiography,myocardial contrast echocardiography has the advantage of providing an accurate comprehensive assessment of cardiacstructure and function and denoting the presence or absence of flow-limiting CAD and the status of myocardial viabilityin a single examination in patients presenting with symptomatic left bundle-branch block.




S.A. Hayat, G. Dwivedi, A. Jacobsen, T.K. Lim, C. Kinsey and R. SeniorRadionuclide Perfusion Imaging for the Detection of Coronary Artery Disease

Perfusion Reserve: Implications for Myocardial Contrast Echocardiography Versus Effects of Left Bundle-Branch Block on Cardiac Structure, Function, Perfusion, and

Print ISSN: 0009-7322. Online ISSN: 1524-4539 Copyright © 2008 American Heart Association, Inc. All rights reserved.

is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Circulation doi: 10.1161/CIRCULATIONAHA.107.726711

2008;117:1832-1841; originally published online March 31, 2008;Circulation.

http://circ.ahajournals.org/content/117/14/1832World Wide Web at:

The online version of this article, along with updated information and services, is located on the

http://circ.ahajournals.org//subscriptions/

is online at: Circulation Information about subscribing to Subscriptions:

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

document. Permissions and Rights Question and Answer this process is available in the

click Request Permissions in the middle column of the Web page under Services. Further information aboutOffice. Once the online version of the published article for which permission is being requested is located,

can be obtained via RightsLink, a service of the Copyright Clearance Center, not the EditorialCirculationin Requests for permissions to reproduce figures, tables, or portions of articles originally publishedPermissions:


http://circ.ahajournals.org/content/117/14/1832

http://www.ahajournals.org/site/rights/

http://www.lww.com/reprints

http://circ.ahajournals.org//subscriptions/


effects of left bundle-branch block on cardiac structure, function, perfusion, and perfusion...

Documents